1. Field of the Invention
The present invention relates to an image compression technique and, more particularly, to an image compression technique to compress an information plane indicating attribute data unique to each pixel of image data of one page.
2. Description of the Related Art
In a conventional image generation apparatus, attribute data is added to each pixel in order to discriminate between texts/graphics/images present on the generated image. However, upon adding attribute data, the memory size required by the image generation apparatus inevitably increases. Hence, when the image generation apparatus has run out of memory, the attribute data is compressed to reduce the required memory size. However, texts and complex graphics, for example, have attribute data complexly inserted in them, so an expected compression result often cannot be obtained. To combat this situation, if attribute data does not fall within a predetermined size, a scheme of discarding this attribute data and setting one attribute data determined for the entire page has been adopted (for example, Japanese Patent Laid-Open No. 2003-209698). Alternatively, a scheme of dividing the entire page into blocks each having a predetermined size, and setting one attribute data determined for each block has been adopted (for example, Japanese Patent Laid-Open No. 2004-134962).
Unfortunately, when all attribute data are held for each pixel, the amount of attribute data becomes relatively large. Also, when one attribute data is set for the entire page or each block, as described above, all parts including texts, graphics, and images undergo uniform image generation processing, thus making it impossible to accomplish the original purpose of use of attribute data.